Previous research has shown that electoral turnout is highly clustered within households. One explanation for this is that as well as sharing common characteristics (such as partisanship and interest in politics) people live together mobilise each other to vote. However, this mobilisation has often thought to be indirect, channelled through the mediating effect of political interest or shared social norms (such as civic duty). However, an equally plausible explanation offered by Fieldhouse and Cutts (2012) is the so called 'companion effect'. This explanation suggests that at least part of the correlation may be attributed to the possibility that the decision to visit the polling station may be made jointly on the day, in the context of other family activities. Until now there has been almost no evidence to support or refute this argument. In this paper we draw on evidence from two unique sources. First we gathered qualitative and quantitative evidence from two polling stations in Manchester (England) observing the groups who visited the polling station and the reasons they gave for voting. Second we commissioned questions in national election studies in Canada, Scotland, Wales and Italy, and a local election study in Oldham (England) which explored who people voted with on polling day.
